The present invention relates to a management system (hereinafter referred to as a manager) for a computer system and more particularly to a manager for monitoring and controlling fault and performance of a plurality of computers on a network through the network in a client server system (hereinafter referred to as a CSS) in which a plurality of computers are connected by the network such as a local area network (hereinafter referred to as a LAN) or a public line.
As the performance of a personal computer and a workstation is enhanced, it has become popular to connect a plurality of computers through a LAN to construct a system. In the CSS, various types of operating systems (hereinafter referred to as network OSs) are mounted on computers functioning as clients or server and these computers are connected through the network to operate in a cooperative manner.
In such a CSS system, as a scale thereof expands, the number of computers functioning as the clients and servers extends to several hundreds to several thousands and when the network manages and operates the respective computers manually by the system operators, the loads of the system operators and the cost therefor increase, which causes a problem. In order to solve such a problem, a management system has been proposed in which a manager is mounted on at least one of the computers on the LAN and an agent for managing each computer in accordance with a direction by the manager is mounted in each of the servers to be managed connected to the LAN so that central management is attained by the manager through the network. Particularly, when a CSS system for a fundamental business of an enterprise is to be constructed, a high reliability is required for the server and a management system focused on fault management of the server is provided.
In such a system focused on the fault management, as disclosed in U.S. Pat. No. 5,367,670, a dedicated extended board for monitoring fault which mounts a processor independent from a main unit is connected to an extended interface of a computer main unit so that fault information collected by the extended board is received by the agent and it is sent to the manager of other management computer connected by the LAN through the network OS to manage the fault. Further, an asynchronous interface is mounted on the extended board so that such collected information may be acquired from the remote computer connected by a modem and a public line and the remote management computer is connected to the extended board via the modem and the public line and the fault is monitored by a dedicated manager. Such remote monitoring offers the advantage of continuous monitoring of fault because the monitoring of fault and the warning are conducted by a processor independent from the main unit on the extended board even if a critical fault which leads to non-operation of the network OS occurs on the computer to be managed (see prior art of FIG. 2).
On the other hand, as a recent trend of the network OS which provides the network environment of the CSS, in order to expand a range of connection of the network and an application area of the system by not only providing the network functions to a plurality of computers connected by the LAN but also connecting mobile computers and office computers, it has been proposed to achieve a remote access function which is totally equivalent to the computers connected by the LAN, between the remote computers connected by the modem and the public line as shown in FIG. 3 to provide a transparent network environment including the public line connection. From the application (AP) operating on the network OS, it is not necessary to know which of the LAN and the line the low order network is connected by, except a communication rate. An example of such network OS is the remote access service (RAS) performed in Windows NT (trade mark) of the Microsoft Inc. US, which is becoming a standard function of the future network OS. Accordingly, the management system for the LAN may be introduced to the line-connected remote computer and uniform management may be attained.
In the above prior art system which is focused on the fault monitoring, no attention is paid to the remote access function of the network OS. Thus, when the system management similar to that of the computer which is LAN-connected to the remote computer connected to the public line by utilizing the remote access function is to be conducted, a remote management computer for managing the system through the agent by utilizing the remote access function, a line and a modem for the connection thereof, a remote management computer directly connected to the fault monitoring extended board for management and a line and a modem for the connection thereof are required. Two management computers, two lines and two modems are required on the management computers and two lines and two modems are required on the computer to be managed. This may cause a problem in the operation. As to the prior art, reference is made to U.S. Pat. No. 5,367,670 to Ward et al and U.S. Pat. No. 5,410,706 to Ferrand et al.